Predict With Observations Plot

Plots predictions of the model with observed values.

Usage

predict_with_observations_plot(
  data,
  fit,
  conc_col,
  dv_col,
  id_col = NULL,
  ntime_col = NULL,
  trt_col = NULL,
  treatment_predictors,
  control_predictors = NULL,
  reference_threshold = c(10),
  conf_int = 0.9,
  contrast_method = c("matched", "group"),
  style = list()
)

Arguments

data

A data frame containing C-QT analysis dataset

fit

An nlme::lme model object from model fitting

conc_col

An unquoted column name for drug concentration measurements

dv_col

An unquoted column name for dependent variable measurements

id_col

An unquoted column name for subject ID

ntime_col

An unquoted column name for nominal time since dose

trt_col

An unquoted column name for treatment group

treatment_predictors

A list for predictions with model. Should contain a value for each predictor in the model.

control_predictors

An optional list for contrast predictions

reference_threshold

Optional vector of numbers to add as horizontal dashed lines

conf_int

Numeric confidence interval level (default: 0.9)

contrast_method

A string specifying contrast method when using control_predictors: "matched" for individual ID+time matching (crossover studies), "group" for group-wise subtraction (parallel studies)

style

A named list of arguments passed to style_plot()

Value

A scatter plot of observations with model prediction line and confidence interval ribbon

Examples

data_proc <- preprocess(cqtkit_data_verapamil)
fit <- fit_prespecified_model(
  data_proc,
  deltaQTCF,
  ID,
  CONC,
  deltaQTCFBL,
  TRTG,
  TAFD,
  "REML",
  TRUE
)
predict_with_observations_plot(
  data_proc,
  fit,
  CONC,
  deltaQTCF,
  treatment_predictors = list(
    CONC = 0,
    TRTG = "Verapamil HCL",
    TAFD = "2 HR",
    deltaQTCFBL = 0
  ),
  conf_int = 0.9
)